MASTER AND PUPIL
Celestial Mechanics of Laplace, thereby laying the foundations
for his lifelong interest in dynamics and systems of simultaneous
differential equations. Of course he could get none of this
through the head of his cultured, petty-official father, and his
obedient brother and his dismayed sisters knew not what the
devil he was talking about. The fact alone was sufficient:
brother Karl, the genius of the timorous little family, on whom
such high hopes of bourgeois respectability had been placed,
had come home, after four years of rigid economy on father's
part, without a degree.
At last - after weeks - a sensible friend of the family who had
sympathized with Karl as a boy, and who had an intelligent
amateur's interest in mathematics, suggested a way out: let
Karl prepare himself at the neighbouring Academy of Minister
for the state teachers' examination. Young Weierstrass would
not get a Ph.D. out of it, but his job as a teacher would provide
a certain amount of evening leisure in which he could keep alive
mathematically provided he had the right stuff in him. Freely
confessing his lsins' to the authorities, Weierstrass begged the
opportunity of making a fresh start. His plea was granted, and
Weierstrass matriculated on 22 May 1839 at Miinster to prepare
himself for a secondary school teaching career. This was a most
important stepping stone to his later mathematical eminence,
although at the time it looked like a total rout.
What made all the difference to Weierstrass was the presence
at Munster of Christof Gudermann (1798-1852) as Professor of
Mathematics. Gudermann at the time (1839) was an enthusiast
for elliptic functions. We recall that Jacobi had published his
Fundamenta nova in 1829. Although few are now familiar with
Gudermann's elaborate investigations (published at the instiga-
tion of Crelle in a series of articles in his Journal), he is not to be
dismissed as contemptuously as it is sometimes fashionable to
do merely because he is outmoded. For his time Gudermann
!iad what appears to have been an original idea. The theory of
illiptic functions can be developed in many different ways - too
nany for comfort. At one time some particular way seems the
>est; at another, a slightly different approach is highly adver*
ised for a season and is generally regarded as being more chic*
455